Search

Sign language has become increasingly popular as a means of language intervention for the mentally handicapped. Much research has centred on various aspects of single sign acquisition, but there is a lack of detailed and systematic research on the acquisition of more complex levels of sign use. The majority of work in this area fails to use or report the specific methods employed. One study, however, by Karlan et al. (1982) did introduce a potentially useful technique, derived from Miniature Linguistic Systems research, that was specifically aimed at establishing the use of generalised, recombinable, two-sign combinations. The aim of the research in this thesis is to investigate aspects of the technique, termed 'matrix training', with special emphasis on areas of generalisation. The research involved a series of five, long-term, school-based teaching programmes using children classified as 'Severe Learning Difficulties'. A single-subject, multiple baseline/probe methodology was employed. Training and testing sessions were alternated in order to follow the emergence of any generalised responses. Each separate study examined specific areas of generalisation associated with the technique. Results indicated that matrix training was successful, within certain conditions, in establishing the use of productive, expressive, two-sign combinations over a heterogenous group of children, and using three differnet forms of combination. There was also evidence for generalisation beyond the original training situation, and of generalisation which enabled the children to incorporate novel elements into their productive responses. Consideration was also given to the children's comprehension of relationships encoded within two-sign combinations in the form of word/sign order information. In the reversible object-location combinations used, it seemed unlikely that the children used sign order to identify object or location items, but rather used their own 'nongrammatical' strategies. Consideration was also given to how, or if, matrix training procedures could be used to establish the use of sign order in this manner. Overall, mentally hadicapped children, who do not typically use expressive sign (or word) combinations, can under certain circumstance be brought to use productive sign combinations by systematic teaching procedures of the kind described. These need to be highly contrived if they are to be effective, and what is learned may be sensitive to the precise characteristics of the matrices used. The potential role of matrix training within language intervention is seen as only a part of a more general sign programme. Discrete-trial matrix training procedures could be used to establish the structure of the two-sign combination, while incidental learning could develop the use of combination within the child's natural environment. Evidence from this research is favourable for this type of combined approach within sign language intervention.